India's AI Med-Tech Push: Growth Engine or Health Equity Playbook?
Last week's AI Summit energy was unmistakable: retinal AI for early blindness, handheld ECGs promising quick risk scores, and ICU dashboards predicting deterioration. The signal is clear-India wants to lead in AI-driven medicine. But behind the demos sits a bigger shift: healthcare is being reframed as an industry, not just a public service.
Through Production Linked Incentives and Make in India, the country is building a med-tech manufacturing base to cut imports, attract capital, and win exports. That ambition is sound, especially after COVID exposed supply chain fragility. India's medical devices market is roughly $11-12 billion, yet 70-80% of high-end equipment is still imported. For a nation this size, that dependence is a strategic risk.
The incentive split: growth vs. health
Industrial policy rewards scale, margins, and exports. Public health cares about affordability, equity, and disease reduction. These goals overlap, but they don't automatically align. Tie incentives to sales value, and companies will chase high-value SKUs, urban hospital chains, and global price points.
This isn't unethical-just how markets work. Markets optimise for solvency; public health optimises for survival. If we don't wire the incentives, we'll get impressive tech that struggles to reach PHCs and district hospitals.
Lessons from elsewhere
China built a powerful med-tech export engine with strong state support, yet needed targeted policy to address rural gaps. Scale didn't deliver equity by itself. The United States leads in devices and AI diagnostics, yet costs remain high and access uneven.
Brazil offers a different pattern: a strong public system (SUS) and procurement that supports domestic production aligned to public need. India uses large public purchasing through Ayushman Bharat and government tenders, but with nearly half of spending still out-of-pocket, the private market heavily shapes what gets built and priced.
India's current baseline
Public health spending sits near 2% of GDP. Out-of-pocket payments remain close to half of total health spend-far higher than the United States (about 10-11%), Brazil (22-25%), China (27-30%), and the United Kingdom (around 15%). Two decades ago, China was where India is today and brought OOP down by expanding insurance, growing public spend, and negotiating prices hard.
Primary care capacity is still uneven across states while capital chases tertiary systems and AI diagnostics. That creates a design bias: build for high-resource settings first, retrofit later. Meanwhile, the biggest burden-undiagnosed hypertension, TB detection gaps, maternal risk, anaemia-sits where margins are thin.
The real question
India should pursue med-tech leadership. The decision to make is whether domestic public health need anchors that ambition. If industrial policy and health policy run on separate tracks, we'll build an advanced but unequal system.
The alternative: design for Bharat first. Build for constraints-price, power, connectivity, workforce-and you create solutions that travel globally because they're proven where it's hardest to operate.
A practical playbook for healthcare leaders
For policymakers and payers
- PLI + Access: Tie a portion of incentives to verifiable domestic deployment in public facilities, rural usability, and capped total cost of ownership (device + disposables + service).
- Procurement scoring: Weight tenders for affordability, PHC readiness (offline mode, battery life, training hours), open standards, and service network density.
- Price-volume + outcome contracts: Negotiate lower unit prices in exchange for scale and performance guarantees (uptime, turnaround time, sensitivity/specificity for AI).
- Coverage first: Fast-track reimbursement under Ayushman Bharat for primary care diagnostics with strong health-economic value, not just tertiary optimisation.
- Data guardrails: Require local validation, explainability summaries, bias audits across languages and demographics, and clear human-in-the-loop workflows.
For hospital systems and state programs
- Buy the outcome, not just the device: Contract on uptime, time-to-result, and referral accuracy. Insist on AMC cost ceilings and local repair SLAs.
- Interoperability or no deal: Prefer vendors that support open APIs, FHIR, and vendor-neutral archives to avoid lock-in.
- Pilot where it's hard: Validate in district hospitals and PHCs before city flagships. If it works there, it will work anywhere.
- Train the last mile: Bundle onboarding hours, competency checklists, and refresher modules. Make training a contractual deliverable.
For startups and device makers
- Design constraints upfront: Build for intermittent power, offline workflows, low-cost consumables, and minimal calibration.
- Unit economics that survive PHCs: Aim for razor-thin recurring costs and price-volume paths that public payers can stomach.
- Evidence that matters: Prospective studies in tier-2/3 settings, local languages, and with ASHAs/ANMs where relevant. Publish or at least pre-register protocols.
- Serviceability as a feature: Modular parts, remote diagnostics, and a distributed repair network beat flashy specs that sit idle.
Metrics that keep everyone honest
- Access: Share of devices deployed in PHCs/CHCs; rural vs. urban split; monthly active usage by facility tier.
- Affordability: Out-of-pocket spend per episode; price per test; TCO over three years vs. baseline.
- Clinical impact: Time-to-diagnosis, treatment initiation rate, and reduction in missed cases (e.g., TB, hypertension, maternal risk).
- Quality and safety: Sensitivity/specificity in local validation, false alarm rates, and audit outcomes for bias and drift.
- Reliability: Uptime, repair turnaround, consumable stock-outs.
Risks to avoid
- Vendor lock-in: Proprietary ecosystems that trap data and inflate lifecycle costs.
- Dataset bias: Models trained on urban, high-resource cohorts performing poorly in rural settings.
- Spec inflation: Overbuilt features that raise price without improving outcomes.
- Hype over service: Great algorithms with weak after-sales support end up gathering dust.
The path forward
Manufacturing strength is necessary. Equity makes it durable. If incentives reward domestic deployment, PHC usability, and real-world outcomes-not just output and exports-India can build a med-tech engine that serves its people first and still wins globally.
The summit showed ambition. The next move is policy design: wire growth to access.
Useful references
- Government of India: PLI Scheme for Medical Devices
- World Bank: Out-of-Pocket Health Expenditure (India)
Want practical how-tos on implementing AI in care pathways?
AI for Healthcare offers hands-on resources for clinicians and health system leaders building safe, effective AI-enabled services.
Your membership also unlocks:
